Dimethyl ether oxidation analyzed in a given flow reactor: Experimental and modeling uncertainties

Dimethyl ether (DME), a widely studied alternative fuel, is known to exhibit complex low- and high-temperature oxidation chemistry. It is also the smallest molecule in the families of symmetric ethers and oxymethylene ethers that receive attention as renewable fuels. Thanks to several studies performed in facilities such as shock tubes, jet-stirred reactors and flames, it can be assumed that the DME oxidation is well understood. However, DME oxidation in flow reactors has been addressed comparatively rarely, although this configuration presents an interesting system with influences of both kinetics and fluid dynamics on the reaction behavior. To examine the interplay of both influences and potential uncertainties resulting from such effects, DME oxidation was experimentally investigated over an extended range of conditions in a flow reactor equipped with mass-spectrometric analysis. Quantitative species profiles were obtained at near-atmospheric pressure in a temperature range of 400–1100 K for three equivalence ratios φ (0.8, 1.0 and 1.2), and three flow rates at each stoichiometry. These nine different cases were first analyzed using a detailed chemical reaction mechanism with a Plug-Flow Reactor (PFR) model. In-depth examination of experimental and reaction model uncertainties led to updates in the reaction mechanism that were performed on the basis of most recent, reliable kinetic information. In spite of the good agreement of the PFR model with the experimental data at selected conditions, especially in the low-temperature regime, substantial deviations in the reactivity and associated species profiles were noted in several cases, particularly for lean conditions at low flow rates and intermediate temperatures around and above 700 K. A two-dimensional (2D) computational fluid dynamics (CFD) model was therefore employed to characterize the reactive flow conditions more accurately. Significant contributions of fluid dynamics effects were observed in the cases that presented the most severe deviations, and overall good agreement within experimental uncertainty was obtained for the nine cases with the 2D simulations. With the aid of a mathematical curve matching procedure using a variety of recent, established kinetic mechanisms, it could be convincingly demonstrated under the current conditions that improvements in predictive modeling capability in the sensitive test cases were not a question of improved kinetics but were mainly achieved by considering the two-dimensional reactive flow. As a consequence, the present investigation can serve to alert the community to the potential major influences that might be neglected if standard PFR models are used to predict fuel oxidation without detailed analysis whether the conditions are suited to that approach.